ERCP in Patients with Prior Billroth II Gastrectomy: Report of 30 Years’ Experience

Original article 611

ERCP in patients with prior Billroth II gastrectomy: report of 30 years’ experience

Authors Vincenzo Bove1, Andrea Tringali1, Pietro Familiari1, Giovanni Gigante1, Ivo Boškoski1, Vincenzo Perri1, Massimiliano Mutignani2, Guido Costamagna1

Institutions 1 Digestive Endoscopy Unit, Catholic University, Gemelli University Hospital, Rome, Italy 2 Digestive Endoscopy Unit, Niguarda Hospital, Milan, Italy

submitted 18. June 2014 Background and study aim: Endoscopic retro- for intubation failure was a long and angulated af- accepted after revision grade cholangiopancreatography (ERCP) is diffi- ferent loop. Successful cannulation/opacification 28. December 2014 cult in patients with altered anatomy following of the desired biliopancreatic duct was 93.8 % Billroth II gastrectomy. Afferent loop intubation, (580/618). Biliary and/or pancreatic sphinc- Bibliography selective cannulation, and sphincterotomy are terotomy were performed in 490 (84.5%) and 23 DOI http://dx.doi.org/ the main issues. Experience from a tertiary refer- (4.0%) patients, respectively. The adverse event 10.1055/s-0034-1391567 ral endoscopy center is reported. rate was 4.3 % (45/1050 procedures). Peritoneal Published online: 2.3.2015 Patients and methods: A total of 713 patients perforation occurred in 1.8 % of the cases (19/ Endoscopy 2015; 47: 611–616 with Billroth II reconstruction who underwent 1050 procedures) and always required immediate © Georg Thieme Verlag KG Stuttgart · New York ERCP between October 1982 and October 2012 surgery. Two patients died after surgery (overall ISSN 0013-726X were retrospectively identified from a prospec- mortality 0.3%). The other adverse events re- tively collected database (mean age 69±27 years; solved following conservative management or Corresponding author 567 males). The main indications for ERCP were endoscopic reintervention. Andrea Tringali, MD Digestive Endoscopy Unit common bile duct stones (51.2%) and obstructive Conclusions: In experienced centers, ERCP in Bill- Catholic University jaundice (24.8%). Procedures were always started roth II patients had morbidity and mortality rates Gemelli University Hospital with a duodenoscope; in cases of failure to reach that were comparable to patients with normal Largo A. Gemelli 8 the papilla the duodenoscope was changed to a anatomy. The main reasons for failure were relat- Rome 00168 gastroscope. Endoscopic sphincterotomy was ed to the inability to reach the papilla. Peritoneal Italy performed using a long-nose sigmoid inverted perforation was the most common adverse event, [email protected] sphincterotome. and required a prompt surgical approach. Results: The successful duodenal intubation rate was 86.7 % (618/713 patients). The main reason

Introduction in patients with Billroth II and normal anatomy !

[10,11]. Downloaded by: IP-Proxy Ospedale G. Salvini, Salvini. Copyrighted material. Endoscopic retrograde cholangiopancreatogra- The aim of this study was to analyze a 30-year ex- phy (ERCP) is considered to be the first-line treat- perience from a tertiary referral center carrying ment several biliary and pancreatic diseases. out ERCP in patients with prior Billroth II gas- ERCP may also be required in patients with surgi- trectomy. cally altered anatomy, for example following Bill- roth II gastrectomy [1]. ERCP in patients with Bill- roth II anatomy is a challenging procedure. Enter- Patients and methods ing the afferent loop, progressing towards the ! duodenal stump, visualizing the papilla, cannu- Patients lating the desired duct (biliary and/or pancreatic), From October 1982 to October 2012, 1050 (4.2%) and performing endoscopic sphincterotomy in a out of a total of 25 000 ERCPs were performed in reverse position, are the main issues [2]. ERCP in 713 patients with previous Billroth II gastrectomy normal anatomy has a 0.1%– 0.6% risk of perfora- at the Digestive Endoscopy Unit of Gemelli Uni- tion [3–7], whereas in Billroth II patients this fig- versity Hospial, Rome, Italy. The indication for ure may increase up to 10.2% [8, 9]. The incidence Billroth II gastrectomy was peptic ulcer in 623 pa- Scan this QR-Code to watch of other ERCP-related adverse events (pancreati- tients (87.4 %) and gastric neoplasia in 90 patients the video comment. tis, bleeding, cholangitis, cholecystitis) is similar (12.6 %). These patients were retrospectively

Bove Vincenzo et al. ERCP in patients with prior Billroth II anatomy… Endoscopy 2015; 47: 611–616 612 Original article

Table1 Characteristics of the 713 patients with Billroth II anatomy.

n% Males 567 79.5 Females 146 20.5 Age≤ 60 years 148 20.8 Age>60 years 565 79.2 Gastric resection for peptic ulcer 623 87.4 Gastric resection for cancer 90 12.6

identified from a prospectively collected database (567 males, 146 females; mean age 69 years [range 27–96]) (●" Table1). This study was approved by the ethical committee of the Catholic University of Rome on 19 April 2012 (P/445/CE/2012).

Study end points The primary end point of the current analysis was the success rate of duodenal stump intubation during ERCP. Secondary end points were: 1) success in cannulation of the desired duct; 2) therapeutic success; 3) evaluation of ERCP-related adverse events and their management.

ERCP procedures ERCP in patients with previous Billroth II gastrectomy was always Fig. 1 Endoscopic retrograde cholangiopancreatography in patients with previous Billroth II gastrectomy. The endoscope reaches the left abdominal approached with a side-viewing operative duodenoscope (Olym- quadrant towards the pelvis, and the efferent loop is intubated. pus TJF100, 140, 160; Pentax ED 3440, ED 3470TK; Fujinon DUO- XL, DUO-XT). In cases of failure to reach the papilla, the duodenoscope was changed to a forward-viewing gastroscope or a pedia- Fig. 2 Retraction of tric colonoscope. The procedure was performed under conscious the endoscope from the sedation with fentanyl and midazolam or deep sedation/general efferent loop with the anesthesia with propofol. catheter in place; the ERCP was started with the patient in the left lateral decubitus po- afferent loop is identi- sition to facilitate entering the afferent loop.After loop intubation fied (arrow). the patient was turned to the supine position in order to obtain better radiological images. The loop anastomized towards the lesser gastric curvature was the first to be intubated because the majority of the afferent loop is in this location. If the intubated loop looked “dry,” without evidence of bile and the endoscope was in the left abdominal quadrant toward the pelvis on radio- graph (●" Fig.1), the other loop was entered (●" Fig. 2). In recent years, if the afferent loop had tight and sharp angulations, a catheter and a soft angled guidewire (Terumo Radiofocus, Tokyo, Ja-

pan) was advanced under radiological control to the duodenal Downloaded by: IP-Proxy Ospedale G. Salvini, Salvini. Copyrighted material. stump to act as a “road map” in order to determine whether it was feasible to proceed with the duodenoscope (●" Fig.3). Cannulation was performed using standard straight ERCP catheters (ERCP-1-HKB, Cook Endoscopy, Winston Salem, North Caro- ing catheters. This guarantees an almost constant reverse posi- lina, USA) or, in more recent years, bendable catheters (Swing Tip, tion of the cutting wire (at 6 o’clock) and allows the possibility Olympus Medical, Tokyo, Japan). Straight catheters are needed in of wire guidance [12] (●" Fig.4). This inverted sphincterotome is Billroth II patients because the direction for cannulation is at the today manufactured by Endobair (London, UK). Other groups 6o’clock position; steerable catheters can facilitate the correct have also designed similar sphincterotomes with good efficacy approach to the papilla in this reverse position. A standard con- [13,14]. trast injection technique was used for cannulation as a first-line Standard ERCP devices were used for other therapeutic proce- approach; if this approach failed, guidewire-assisted cannulation dures (e.g. stone extraction, stent insertion). was the second approach, and needle-knife precut was the third. In cases of multiple common bile duct stones and doubtful com- Endoscopic sphincterotomy in Billroth II patients is performed in plete stone clearance, a nasobiliary drain (NBD) was placed in or- a reverse position. In the current series, endoscopic sphincterot- der to perform a repeat cholangiography to check the area and to omy was initially performed with the Cremer-Van Laethem in- facilitate repeat ERCP if needed. verted sphincterotome. From 1987, a long-nose “homemade” sigmoid inverted sphincterotome was obtained by modifying exist-

Bove Vincenzo et al. ERCP in patients with prior Billroth II anatomy… Endoscopy 2015; 47: 611–616 Original article 613

Fig. 4 The sphincterotome enables controlled sphincterotomy. a The “homemade” sigmoid inverted sphincterotome. b The sphincterotome is wire guided. c The cutting wire is directed at the 6o’clock position.

Fig. 3 A catheter is advanced into the duodenal stump.a The afferent loop is short and not angled. b A long and angled afferent loop; the procedure was interrupted.

At the end of the procedure the site of the afferent loop (lesser or

greater curvature) was annotated in the database to facilitate fu- Downloaded by: IP-Proxy Ospedale G. Salvini, Salvini. Copyrighted material. ture reinterventions.

Key technical aspects of ERCP in Billroth II patients Table2 Indications for endoscopic retrograde cholangiopancreatography ▶ Begin ERCP in Billroth II patients with a therapeutic duodeno- in 713 patients with prior Billroth II gastrectomy. scope; switch to a forward-viewing endoscope if the papilla cannot be reached. n% ▶ Place the patient in the left lateral decubitus position and turn Common bile duct stones 365 51.2 them to the supine position after afferent loop intubation. Obstructive jaundice 177 24.8 ▶ Use fluoroscopy to assess correct loop intubation. Acute cholangitis 61 8.6 1 ▶ Use an inverted sphincterotome. Chronic pancreatitis 55 7.7 Acute biliary pancreatitis 21 2.9 External biliary fistula 20 2.8 Benign biliary strictures2 91.3 Results External Pancreatic fistula 3 0.4 ! Duodenal fistula3 20.3 A total of 1050 ERCPs were attempted in 713 patients with pre- 1 Six also underwent extracorporeal shock wave lithotripsy. vious Billroth II gastrectomy (1.47/patient; range 1–33). The in- 2 Eight postoperative biliary stricture, 1 hydatid cyst-related biliary stricture. dications for ERCP are summarized in●" Table 2. 3 Nasobiliary and nasopancreatic drains were inserted.

Bove Vincenzo et al. ERCP in patients with prior Billroth II anatomy… Endoscopy 2015; 47: 611–616 614 Original article

Fig. 5 Endoscopic retrograde cholangiopancrea- 60 tography failure in patients with Billroth II anatomy (n=713) during 30 years of experience. 50

30 Failure rate (%) rate Failure 20

0 Oct 1982 Nov 1987 Nov 1992 Nov 1997 Nov 2002 Nov 2007 Oct 1987 Oct 1992 Oct 1997 Oct 2002 Oct 2007 Oct 2012 n = 26/48 n = 21/131 n = 17/121 n = 20/161 n = 25/143 n = 24/109

Failure (%) 54.1 16 14 12.4 17.5 22

Table3 Results of duodenal intubation and biliopancreatic cannulation Table5 Endoscopic retrograde cholangiopancreatography in patients with after endoscopic retrograde cholangiopancreatography in 713 patients with Billroth II anatomy: adverse events in 1050 procedures. Billroth II anatomy. n Specific, % Overall, % n% Morbidity 45 4.3 Successful duodenal intubation 618 86.7 Peritoneal perforation 19 42.3 1.8 Duodenoscope 600 97.1 Delayed postendoscopic 11 24.4 1.0 Gastroscope 11 1.8 sphincterotomy bleeding Pediatric colonoscope 7 1.1 ERCP-related cholangitis 5 11.1 0.5 Failed duodenal intubation 95 13.3 Mild pancreatitis 5 11.1 0.5 Long and angulated afferent loop 73 76.8 Retroperitoneal perforation 3 6.7 0.3 Jejunal loop perforation 19 20.0 Respiratory failure 2 4.4 0.2 Lack of cooperation (conscious sedation) 3 3.2 ERCP, endoscopic retrograde cholangiopancreatography. Successful cannulation/opacification 580/618 93.8 Duodenoscope 567/600 94.5 Gastroscope 13/18 72.2 of experience and remained stable during the subsequent years Overall success 580/713 81.3 (●" Fig.5). Duodenal intubation failed in 95 patients (13.3%). The reasons for failure were: long and angulated afferent loop, including patients with Braun side-to-side jejuno-jejunostomy (n Table4 Endoscopic retrograde cholangiopancreatography procedures on =73 [76.8%]); jejunal loop perforation (n=19 [20%]); and insuffi- 580 patients with previous Billroth II gastrectomy. cient sedation and/or lack of cooperation (n=3 [3.2%]). The total

n%number of successful opacification and cannulation procedures was 580/618 (93.8 %; overall success rate 81.3 %) (●" Table 3). Biliary sphincterotomy1 490 84.5 In this series, 500 patients out of 580 (86.2%) required therapeu- Extraction of common bile duct stones 318 54.8 "

tic interventions (● Table 4). Among 365 patients with common Downloaded by: IP-Proxy Ospedale G. Salvini, Salvini. Copyrighted material. Plastic and metallic biliopancreatic stents 158 27.2 bile duct stones, 100 (27.4%) received an NBD because multiple Diagnostic ERCP2 78 13.4 Pancreatic sphincterotomy3 23 4.0 stones had been present. Repeat cholangiography showed resi- Balloon dilation of a previous sphincterotomy4 19 3.3 dual fragments to be present in 20 of these cases. Repeat ERCP Extraction of pancreatic stones/plugs5 12 2.1 in these patients was always successful without adverse events. Papillectomy 1 0.2 The overall morbidity in 1050 procedures was 4.3 % (n=45; 95% – " ERCP, endoscopic retrograde cholangiopancreatography. CI 3.1 5.5) (● Table 5). The most common adverse event of the 1 Including 53 (10.8%) needle-knife precut. procedure was peritoneal perforation in 19 patients (19/713 pa- 2 Diagnostic procedures were performed when magnetic resonance cholangio- tients [2.7%]; 19/1050 procedures [1.8%]), which was recognized pancreatography was not yet available. 3 Including 4 minor papilla sphincterotomy; 16 patients also had biliary sphinctero- immediately during endoscopy and always treated by urgent sur- tomy. gery. No attempt was made to repair or mark the perforation site 4 Surgical or endoscopic. endoscopically. During surgery the perforation site was easily 5 Six patients underwent extracorporeal shock wave lithotripsy. identified. All of the perforations occurred when using the duodenoscope to intubate the afferent loop and were less frequent The duodenal stump was successfully reached in 618 patients in patients under conscious sedation with midazolam (15/1012; (86.7 %; 95 % confidence interval [CI] 84.2 –89.2) –600 with a 1.4 %) than in those under general anesthesia (4/38, 10.5 %; P< duodenoscope, 11 with a gastroscope, and 7 with a pediatric co- 0.01 [chi-squared test]). lonoscope. The overall failure rate decreased after the first 5 years

Bove Vincenzo et al. ERCP in patients with prior Billroth II anatomy… Endoscopy 2015; 47: 611–616 Original article 615

Table6 Results from published series on endoscopic retrograde cholangiopancreatography in patients with Billroth II anatomy.

Study Patients, Type of Successful afferent Successful Afferent loop Mortality due First author, n endoscope loop intubation, % biliopancreatic peritoneal to peritoneal year [ref.] cannulation, % perforation, % perforation, % Osnes, 1986 [2] 147 Side viewing 92 100 0.7 0.7 Costamagna, 1994 175 Side viewing 94.7 93.2 1.1 0 [12] Hintze, 1997 [13] 59 Side viewing 92 100 N/A N/A Kim, 1997 [9] 23 Forward 91.3 95.2 0 0 viewing Kim, 1997 [9] 22 Side viewing 68 100 18.2 0 Lin, 1999 [18] 56 Forward 76.8 81.4 0 0 viewing Çiçek, 2007 [8] 59 Side viewing 86.4 88.2 10.2 1.7 Nakahara, 2009 43 Prototype anterior 88.3 94.7 0 0 [19] oblique viewing Current series 713 Side viewing 86.7 93.8 2.7 0.3 N/A, not addressed.

Peritoneal perforation during the first ERCP occurred in 17/713 successful with the side-viewing endoscope. The rate of success- patients (2.4%; 95 %CI 1.3– 3.5), and 2/337 patients (0.6 %; 95%CI ful sphincterotomy was comparable between the two groups (83 0.1 –1.4) experienced perforation during the subsequent ERCPs % vs. 80 %, respectively). Another possible advantage of the for- performed for biliary stent exchange (P =0.0294 [chi-squared ward-viewing endoscope was a lower rate of adverse events. test]). Kim et al. reported jejunal perforation in 4 of 22 patients (18.2%) Two patients died after surgery (overall mortality 2 /713 [0.3%]; using the side-viewing endoscope compared with none with the specific mortality 2 /19 [10.5%]). Other early adverse events in- forward-viewing endoscope. cluded postsphincterotomy bleeding, retroperitoneal perfora- The reported success in therapeutic ERCP in patients with pre- tions, ERCP-related cholangitis (3 malfunctioning biliary stent, 2 vious Billroth II gastrectomy is slightly higher with the duodeno- residual stones), mild pancreatitis, and respiratory failure (●" Ta- scope than the forward-viewing endoscope (●" Table 6). Accord- ble5). These adverse events resolved by conservative manage- ing to the authors’ experience, the therapeutic duodenoscope is ment or endoscopic reintervention. The three retroperitoneal preferred, making biliopancreatic cannulation and subsequent perforations were secondary to sphincterotomy, balloon dilation endotherapy (even pancreatic) easier thanks to better visualiza- of the papilla, and laceration of a duodenal diverticulum, respec- tion of the papilla, the presence of the elevator, and the large op- tively, and all were managed conservatively. erative channel. In recent years, the introduction of balloon-assisted enteroscopy has increased the success rate of duodenal stump intubation Discussion when performing ERCP in Billroth II patients [16]. Nevertheless, ! the small operative channel of the enteroscope (2.8mm), the for- ERCP in patients who have undergone previous Billroth II gas- ward view, the lack of the elevator, and the absence of ERCP-dedi- trectomy is a challenging procedure. Recognition of the afferent cated catheters are limitations that can reduce the therapeutic ef- loop at the site of gastrojejunostomy is the first problem to be ficacy of the enteroscope when performing ERCP in altered anat- solved, but the progression into the loop may then be hindered omy. by situations such as anastomosis angulations, adhesions, exces- However, most published series have confirmed that the duode-

sive length, Braun entero-entero-anastomosis, and distorted noscope presents a higher incidence of perforations due to the Downloaded by: IP-Proxy Ospedale G. Salvini, Salvini. Copyrighted material. anatomy at the duodeno-jejunal angle [12] natural shape of the instruments and the side-viewing character- The choice between a forward-viewing and side-viewing endo- istics [9]. In our experience, the majority of perforations were scope is a matter of debate. Some authors advocate the routine caused by the tip of the endoscope, when excessive scope torsion use of forward-viewing endoscopes (gastroscope or pediatric co- was applied, and when limb visualization was limited. lonoscopies) to manage long afferent loops because they are easi- Advancing a catheter to depict a fluoroscopic “road map” can be er to handle and can facilitate recognition of the afferent loop. useful to pass the angle of Treitz, which is fixed and is the site The major drawbacks of forward-viewing endoscopes are the dif- where perforations usually occur. It is also useful for the identifi- ficult visualization of the papilla and the lack of the elevator cation of anatomical features (i.e. angulations, long loop) in which facilitates cannulation maneuvers [15]. This limitation un- which termination of the procedure can reduce the risk of per- derlies the recommendation of other authors who routinely use foration. side-viewing duodenoscopes, believing that, with experience, Evolution in scope design has reduced the risk of jejunal perfora- the ease of identification of the afferent loop and papilla is com- tion with the duodenoscope in Billroth II patients. The only jeju- parable to the forward-viewing endoscope [15]. nal perforation reported in another series [17] occurred with a Kim at al. [9] randomized 45 patients with a Billroth II gastrect- duodenoscope with a long and potentially harmful distal end omy who underwent ERCP using either a forward- or side-view- (TJF V10, Olympus); this device was very quickly replaced by a ing endoscope. After reaching the papilla, cannulation failed in duodenoscope with a smooth, regular, distal end (TJF 100, Olym- one patient in the forward-viewing group, whereas it was always pus).

Bove Vincenzo et al. ERCP in patients with prior Billroth II anatomy… Endoscopy 2015; 47: 611–616 616 Original article

In the current series, sphincterotomy was performed using an in- References verted sphincterotome. An alternative technique for endoscopic 1 Faylona JM, Qadir A, Chan AC et al. Small-bowel perforations related to sphincterotomy in Billroth II patients is the use of a needle-knife endoscopic retrograde cholangiopancreatography (ERCP) in patients with Billroth II gastrectomy. Endoscopy 1999; 31: 546–549 over a 7-Fr stent. The needle-knife technique is a “free hand” 2 Osnes M, Rosseland AR, Aabakken L. Endoscopic retrograde cholangiog- rather than a well-controlled cutting technique. The wire-guided raphy and endoscopic papillotomy in patients with a previous Billroth- inverted sphincterotome can perform a more controlled endo- II resection. Gut 1986; 27: 1193 –1198 scopic sphincterotomy, giving us more confidence in this partic- 3 Freeman ML, Nelson DB, Sherman S et al. Complications of endoscopic – ular and difficult setting. biliary sphincterotomy. N Engl J Med 1996; 335: 909 918 4 Masci E, Toti G, Mariani A et al. Complications of diagnostic and thera- Leaving an NBD in place in cases of suspected incomplete stone peutic ERCP: a prospective multicenter study. Am J Gastroenterol clearance can facilitate repeat ERCP if necessary. Following the 2001; 96: 417–423 route of the NBD resulted in easier duodenal stump intubation, 5 Loperfido S, Angelini G, Benedetti G et al. Major early complications without perforations, according to our experience. from diagnostic and therapeutic ERCP: a prospective multicenter study. Gastrointest Endosc 1998; 48: 1–10 In the current series, general anesthesia or deep sedation with 6 Colton JB, Curran CC. Quality indicators, including complications, of propofol reduced the rate of procedure failure from lack of coop- ERCP in a community setting: a prospective study. Gastrointest Endosc eration or desaturation, but a significantly (P<0.01) higher rate of 2009; 70: 457–467 perforation occurred compared with procedures performed un- 7 Howard TJ, Tan T, Lehman GA et al. Classification and management of der conscious sedation with midazolam. This fact can be related perforations complicating endoscopic sphincterotomy. Surgery 1999; 126: 658–663 to an earlier procedure termination when the patient reports 8 Ciçek B, Parlak E, Dişibeyaz S et al. Endoscopic retrograde cholangio- pain due to excessive traction on the mesentery. pancreatography in patients with Billroth II gastroenterostomy. J Gas- According to our data, the risk of peritoneal perforation during troenterol Hepatol 2007; 22: 1210–1213 repeated ERCP in the same Billroth II patient, is still present but 9 Kim MH, Lee SK, Lee MH et al. Endoscopic retrograde cholangiopan- is reduced compared with the first attempt. The persistence of creatography and needle-knife sphincterotomy in patients with Bill- roth II gastrectomy: a comparative study of the forward-viewing endo- this risk can be related to different performance of the endoscope scope and the side-viewing duodenoscope. Endoscopy 1997; 29: 82– and to operator skill. 85 A learning curve for advanced ERCP in Billroth II patients is pres- 10 Cohen SA, Siegel JH, Kasmin FE. Complications of diagnostic and thera- – ent, and success increase after the first 50 cases (●" Fig. 5). In our peutic ERCP. Abdom Imaging 1996; 21: 385 394 30-year experience, there has been a slight increase in failures in 11 Huibregtse K. Complications of endoscopic sphincterotomy and their prevention. N Engl J Med 1996; 335: 961–963 more recent years, which may be due to the referral of difficult 12 Costamagna G, Mutignani M, Perri V et al. Diagnostic and therapeutic cases that had already failed in other hospitals. ERCP in patients with Billroth II gastrectomy. Acta Gastroenterol Belg ERCP in altered anatomy is a challenging procedure. The increas- 1994; 57: 155–162 ing number of bariatric surgical procedures, such as Roux-en-Y 13 Hintze RE, Veltzke W, Adler A et al. Endoscopic sphincterotomy using an S-shaped sphincterotome in patients with a Billroth II or Roux-en-Y reconstruction, has led to a growing interest in enteroscopy tech- gastrojejunostomy. Endoscopy 1997; 29: 74–78 niques. Billroth II patients are becoming rare but are still en- 14 Wang YG, Binmoeller KF, Seifert H et al. A new guide wire papillotome countered. for patients with Billroth II gastrectomy. Endoscopy 1996; 28: 254– According to our experience the key technical aspects specific to 255 the Billroth II situation are: to begin with a duodenoscope and to 15 Costamagna G, Loperfido S, Familiari P. Endoscopic retrograde cholangiopancreatography (ERCP) after Billroth II reconstruction. Howell DA, switch to a forward-viewing endoscope in cases of failure to Travis AC, eds. UpToDate Available from: http://www.uptodate.com/ reach the papilla; to position the patient in the left lateral decubi- contents/endoscopic-retrograde-cholangiopancreatography-ercp- tus position and to turn them to the supine position after afferent after-billroth-ii-reconstruction 2014 loop intubation; and to use fluoroscopy to confirm correct loop 16 Shimatani M, Matsushita M, Takaoka M et al. Effective “short” double- intubation, contrast-based cannulation with straight catheters, balloon enteroscope for diagnostic and therapeutic ERCP in patients with altered gastrointestinal anatomy: a large case series. Endoscopy and an inverted sphincterotome for sphincterotomy. 2009; 41: 849–854 17 Demarquay JF, Dumas R, Buckley MJ et al. Endoscopic retrograde cho- Competing interests: Dr. Costamagna has the following declara- langiopancreatography in patients with Billroth II gastrectomy. Ital J Gastroenterol Hepatol 1998; 30: 297–300

tions: Olympus Japan (grant/research support), Cook Inc. (advi- Downloaded by: IP-Proxy Ospedale G. Salvini, Salvini. Copyrighted material. 18 Lin LF, Siauw CP, Ho KS et al. ERCP in post-Billroth II gastrectomy pa- sory committees or review panels; grant/research support), Bos- tients: emphasis on technique. Am J Gastroenterol 1999; 94: 144–148 ton Scientific Corp.(advisory committees or review panels; 19 Nakahara K, Horaguchi J, Fujita N et al. Therapeutic endoscopic retro- speaking and teaching), Given Imaging (speaking and teaching), grade cholangiopancreatography using an anterior oblique-viewing Taewoong Medical Inc. (advisory committees or review panels). endoscope for bile duct stones in patients with prior Billroth II gas- – Dr. Tringali is a former consultant for Boston Scientific. Dr. Boš- trectomy. J Gastroenterol 2009; 44: 212 217 koski is a consultant for Cook Inc.

Bove Vincenzo et al. ERCP in patients with prior Billroth II anatomy… Endoscopy 2015; 47: 611–616